Supplementary material for my lecture on model fitting.
I recommend using anaconda as your Python distribution and package manager. (If you don't already use it, you can download it and install a new version of Python pretty harmlessly. It will install everything to a new directory so it won't mess with your existing installation). One of the great advantages of Anaconda is the ability to create conda environments that let you maintain different, independent Python installations (e.g., you can have Python 2 and 3 installed in parallel) and easily switch between them. I recommend creating a new environment for this material with all of the packages we'll need, just to sandbox everything and make sure we're all using the same versions. If you clone or download this repo, you can use the environment.yml file to automatically create the environment for you. After cloning / downloading, change directories into the project root and do:
conda env create
Once that's done installing everything, you need to activate the environment to use the version of Python with all of the packages it just installed. Do that with:
source activate statsseminar
To deactivate the environment (if you want to go back to your root environment / default context) you just have to do:
source deactivate